Use Case and High-Level Description¶
This is a person, vehicle, bike detector that is based on MobileNetV2 backbone with ATSS head for 448x256 resolution.
AP @ [ IoU=0.50:0.95 ]
0.274 (internal test set)
Average Precision (AP) is defined as an area under the precision/recall curve.
1, 3, 256, 448 in the format
B, C, H, W, where:
B- batch size
C- number of channels
H- image height
W- image width
Expected color order is
boxesis a blob with the shape
100, 5in the format
N, 5, where
Nis the number of detected bounding boxes. For each detection, the description has the format: [
y_min) - coordinates of the top left bounding box corner
y_max) - coordinates of the bottom right bounding box corner
conf- confidence for the predicted class
labelsis a blob with the shape
100in the format
Nis the number of detected bounding boxes. The value of each label is equal to predicted class ID (0 - vehicle, 1 - person, 2 - non-vehicle).
The OpenVINO Training Extensions provide a training pipeline, allowing to fine-tune the model on custom dataset.
The model can be used in the following demos provided by the Open Model Zoo to show its capabilities:
[*] Other names and brands may be claimed as the property of others.